Routine Pathology Research Articles

Initial Application of Fluorescence Imaging for Intraoperative Localization of Small Neuroendocrine Tumors in the Pancreas: Case Report and Review of the Literature.

Indocyanine green is commonly used for laparoscopic hepatectomy but remains uncommon in pancreatic surgery. Given the increasing number of small neuroendocrine tumors found in the pancreas and the heavy reliance on laparoscopic ultrasound for intraoperative localization, we attempted to use indocyanine green for these tumors. Our results show good localization and have the potential to provide a valuable clinical aid. This case report details five patients with preoperative diagnosis of pancreatic neuroendocrine tumors of small endocrine tumors, intraoperative successful localization, and successful completion of laparoscopic partial resection of pancreatic tumors by indocyanine green fluorescence staining; none of the patients experienced serious complications after surgery and were discharged from the hospital, and routine pathology confirmed that four cases were pancreatic neuroendocrine tumors of G1 stage, and one case was pancreatic neuroendocrine cell hyperplasia. Fluorescence imaging technology safely aids in the intraoperative localization of small pancreatic neuroendocrine tumors.

Cancer-Associated Lymphoid Aggregates in Histology Images: Manual and Deep Learning-Based Quantification Approaches.

Quantification of lymphoid aggregates including tertiary lymphoid structures (TLS) with germinal centers in histology images of cancer is a promising approach for developing prognostic and predictive tissue biomarkers. In this article, we provide recommendations for identifying lymphoid aggregates in tissue sections from routine pathology workflows such as hematoxylin and eosin staining. To overcome the intrinsic variability associated with manual image analysis (such as subjective decision-making, attention span), we recently developed a deep learning-based algorithm called HookNet-TLS to detect lymphoid aggregates and germinal centers in various tissues. Here, we additionally provide a guideline for using manually annotated images for training and implementing HookNet-TLS for automated and objective quantification of lymphoid aggregates in various cancer types.

Clinical Pathology Evaluation in Pet Rabbits Vaccinated Against Rabbit Hemorrhagic Disease Virus 2 (RHDV2).

A recombinant vaccine for rabbit hemorrhagic disease virus 2, a highly pathogenic virus, was granted emergency use authorization in the United States after the detection and spread of the virus starting in 2018. The goal of the current study was to assess pet rabbits (n = 29) through physical examination and routine clinical pathology testing using repeated assessments post-vaccination. In addition, seroconversion was also monitored after the initial vaccination and booster vaccination. Neither owners nor clinicians detected any physical abnormalities in relationship to the vaccine protocol. Hematological and clinical biochemistry testing showed some changes although median values were within species specific reference intervals. A significant increase in antibody levels was observed at day 21 (post-initial vaccination) and day 49 (post-booster vaccination) versus that present at baseline (p < 0.0001). However, variability in study rabbits was noted with some individuals showing low antibody levels as well as a lower overall response in older rabbits (r = -0.56, p = 0.006). A second cohort of rabbits was assessed at 11-12 months post-initial vaccination. In this second group, antibody levels were not significantly different from baseline levels (p = 0.21). Additional studies should be conducted to further define the variability in seroconversion and the term of protection in pet rabbits as the industry moves forward in the optimization of RHDV2 vaccines.

A deep learning approach to case prioritisation of colorectal biopsies.

To create and validate a weakly supervised artificial intelligence (AI) model for detection of abnormal colorectal histology, including dysplasia and cancer, and prioritise biopsies according to clinical significance (severity of diagnosis). Triagnexia Colorectal, a weakly supervised deep learning model, was developed for the classification of colorectal samples from haematoxylin and eosin (H&E)-stained whole slide images. The model was trained on 24 983 digitised images and assessed by multiple pathologists in a simulated digital pathology environment. The AI application was implemented as part of a point and click graphical user interface to streamline decision-making. Pathologists assessed the accuracy of the AI tool, its value, ease of use and integration into the digital pathology workflow. Validation of the model was conducted on two cohorts: the first, on 100 single-slide cases, achieved micro-average model specificity of 0.984, micro-average model sensitivity of 0.949 and micro-average model F1 score of 0.949 across all classes. A secondary multi-institutional validation cohort, of 101 single-slide cases, achieved micro-average model specificity of 0.978, micro-average model sensitivity of 0.931 and micro-average model F1 score of 0.931 across all classes. Pathologists reflected their positive impressions on the overall accuracy of the AI in detecting colorectal pathology abnormalities. We have developed a high-performing colorectal biopsy AI triage model that can be integrated into a routine digital pathology workflow to assist pathologists in prioritising cases and identifying cases with dysplasia/cancer versus non-neoplastic biopsies.

New clinicopathological concept of endometrial carcinoma with integration of histological features and molecular profiles.

The dual-stratified pathway of endometrial carcinomas (ECs) has long been dominant. However, in 2013, The Cancer Genome Atlas (TCGA) defined four EC subgroups with distinctive prognoses. Inspired by TCGA, in 2018, the Proactive Molecular Risk Classifier for Endometrial Cancer (ProMisE) provided four pragmatic molecular classifiers to apply surrogate immunohistochemical markers to TCGA subgroup categorization. These trends prompted the revision of 2020 WHO Classification of Female Genital Tumors, 5th edition (2020 WHO classification), in which four molecular subtypes are recognized: POLE-ultramutated; mismatch repair-deficient; p53-mutant; and no specific molecular profile. In the 2020 WHO classification, the diagnostic algorithm is characterized by prioritizing POLEmut over other molecular abnormalities. Following the 2020 WHO classification, Federation of International Gynecology and Obstetrics (FIGO) proposed a new staging system in 2023. The updated system focuses on diagnostic parameters, such as histological type and grade, lymphovascular space invasion, and molecular alterations. These new histomolecular diagnostic concepts of ECs are being accordingly introduced into the routine pathology practice. For the first time, the 2020 WHO classification includes mesonephric-like adenocarcinoma (MLA) as a novel histological entity, mimicking the conventional mesonephric adenocarcinoma, but is considered of Müllerian ductal origin.

Review of In Situ Hybridization (ISH) Stain Images Using Computational Techniques.

Recent advancements in medical imaging have greatly enhanced the application of computational techniques in digital pathology, particularly for the classification of breast cancer using in situ hybridization (ISH) imaging. HER2 amplification, a key prognostic marker in 20-25% of breast cancers, can be assessed through alterations in gene copy number or protein expression. However, challenges persist due to the heterogeneity of nuclear regions and complexities in cancer biomarker detection. This review examines semi-automated and fully automated computational methods for analyzing ISH images with a focus on HER2 gene amplification. Literature from 1997 to 2023 is analyzed, emphasizing silver-enhanced in situ hybridization (SISH) and its integration with image processing and machine learning techniques. Both conventional machine learning approaches and recent advances in deep learning are compared. The review reveals that automated ISH analysis in combination with bright-field microscopy provides a cost-effective and scalable solution for routine pathology. The integration of deep learning techniques shows promise in improving accuracy over conventional methods, although there are limitations related to data variability and computational demands. Automated ISH analysis can reduce manual labor and increase diagnostic accuracy. Future research should focus on refining these computational methods, particularly in handling the complex nature of HER2 status evaluation, and integrate best practices to further enhance clinical adoption of these techniques.

The ATOM-Seq sequence capture panel can accurately predict microsatellite instability status in formalin-fixed tumour samples, alongside routine gene mutation testing

Microsatellite instability (MSI) occurs across a number of cancers and is associated with different clinical characteristics when compared to microsatellite stable (MSS) cancers. As MSI cancers have different characteristics, routine MSI testing is now recommended for a number of cancer types including colorectal cancer (CRC). Using gene panels for sequencing of known cancer mutations is routinely performed to guide treatment decisions. By adding a number of MSI regions to a small gene panel, the efficacy of simultaneous MSI detection in a series of CRCs was tested. Tumour DNA from formalin-fixed, paraffin-embedded (FFPE) tumours was sequenced using a 23-gene panel kit (ATOM-Seq) provided by GeneFirst. The mismatch repair (MMR) status was obtained for each patient from their routine pathology reports, and compared to MSI predictions from the sequencing data. By testing 29 microsatellite regions in 335 samples the MSI status was correctly classified in 314/319 samples (98.4% concordance), with sixteen failures. By reducing the number of regions in silico, comparable performance could be reached with as few as eight MSI marker positions. This test represents a quick, and accurate means of determining MSI status in FFPE CRC samples, as part of a routine gene mutation assay, and can easily be incorporated into a research or diagnostic setting. This could replace separate mutation and MSI tests with no loss of accuracy, thus improving testing efficiency.

Single-cell AI-based detection and prognostic and predictive value of DNA mismatch repair deficiency in colorectal cancer

Testing for DNA mismatch repair deficiency (MMRd) is recommended for all colorectal cancers (CRCs). Automating this would enable precision medicine, particularly if providing information on etiology not captured by deep learning (DL) methods. We present AIMMeR, an AI-based method for determination of mismatch repair (MMR) protein expression at a single-cell level in routine pathology samples. AIMMeR shows an area under the receiver-operator curve (AUROC) of 0.98, and specificity of ≥75% at 98% sensitivity against pathologist ground truth in stage II/III in two trial cohorts, with positive predictive value of ≥98% for the commonest pattern of somatic MMRd. Lower agreement with microsatellite instability (MSI) testing (AUROC 0.86) reflects discordance between MMR and MSI PCR rather than AIMMeR misclassification. Analysis of the SCOT trial confirms MMRd prognostic value in oxaliplatin-treated patients; while MMRd does not predict differential benefit of chemotherapy duration, it correlates with difference in relapse by regimen (PInteraction= 0.04). AIMMeR may help reduce pathologist workload and streamline diagnostics in CRC.

Indirect reference interval estimation using a convolutional neural network with application to cancer antigen 125

Indirect methods for reference interval (RI) estimation, which use data acquired from routine pathology testing, have the potential to accelerate the establishment of RIs to account for variables such as gender and age to improve clinical assessments. However, they require more sophisticated methods of analysis due to the potential influence of pathological patients in raw clinical datasets. In this paper we develop a novel convolutional neural network (CNN) model trained on synthetic data to identify underlying healthy distributions within pathological admixtures. We present both the methodology to generate synthetic data and the CNN model. We evaluate the CNN using two synthetic test datasets, including samples from a proposed benchmark for indirect methods (RIBench) and show significant improvements compared to the reported state-of-the-art method based on the benchmark (refineR). We also demonstrate a real-world application of the model, estimating age-specific RIs for cancer antigen 125 (CA-125), a crucial biomarker for ovarian cancer diagnostics. Our results show that CA-125 RIs are strongly age-dependent, which could have important diagnostic consequences.

Feasibility of Using Needle Rinse Fluid for Cobas Human Papillomavirus (HPV) Assay in Diagnosing HPV+ Oropharyngeal Cancer with Neck Lymph Node Aspiration.

Human papillomavirus (HPV) is a crucial prognostic factor in oropharyngeal cancer (OPC). p16 is a surrogate marker for diagnosing HPV+ OPC, however it is not direct evidence of HPV existence. The purpose of our study was to evaluate an HPV DNA test-Cobas HPV assay-in diagnosing HPV+ OPC through neck lymph node aspiration. Patients with suspected neck mass who received fine needle aspiration (FNA) or core needle biopsy (CNB) at the National Taiwan University Hospital between January 2018 and December 2022 were reviewed. Besides routine cytology and pathology study, needle rinse fluid was collected for the Cobas HPV assay to detect high-risk HPV. We analyzed 137 patients with suspected lymph nodes, 32 (23.4%) of whom were HPV+ OPC patients and 105 (76.6%) of whom had non-HPV-related disease. FNA was performed in 31 patients and CNB was performed in 106 patients, according to the size and necrosis status of the lymph nodes. For diagnosing HPV+ OPC, CNB combined with p16 immunohistochemistry staining showed sensitivity of 93.3%, specificity of 97.8%, positive predictive value (PPV) of 87.5%, negative predictive value (NPV) of 98.9%, and accuracy of 97.2%. On the other hand, for the needle rinse Roche Cobas HPV assay, the test showed sensitivity of 96.9%, specificity of 100%, PPV of 100%, NPV of 99.1%, and accuracy of 99.3%. Compared with p16 IHC staining, the Cobas HPV test showed better PPV with statistical significance (p=0.04). The Cobas HPV assay is a US FDA-approved, highly automated, and readily used technique to directly detect the presence of high-risk HPV. We recommend utilizing the Cobas HPV assay in combination with routine cytology or histopathology examination in the work-up of neck lymphadenopathy.

Endometrial carcinoma: molecular classification in routine pathology

The molecular classification of endometrial carcinoma defines four main groups: polymerase‑ɛ(PolE) gene mutated, microsatellite unstable (MSI), p53 abnormal tumors and tumors with no specific molecular profile (NSMP). This classification provides significant insights into the prognosis and therapeutic decisions. Each group exhibits unique genetic profiles identified through immunohistochemistry and molecular diagnostics, enabling personalized treatment. The identification of these molecular signatures necessitates precise analytical methods, selected based on the local circumstances at each site. The approach to molecular classification highlights the critical role of pathology in the diagnosis and emphasizes the necessity of collaboration between the clinic and pathology.

Artificial Intelligence Models for the Detection of Microsatellite Instability from Whole-Slide Imaging of Colorectal Cancer.

With the advent of whole-slide imaging (WSI), a technology that can digitally scan whole slides in high resolution, pathology is undergoing a digital revolution. Detecting microsatellite instability (MSI) in colorectal cancer is crucial for proper treatment, as it identifies patients responsible for immunotherapy. Even though universal testing for MSI is recommended, particularly in patients affected by colorectal cancer (CRC), many patients remain untested, and they reside mainly in low-income countries. A critical need exists for accessible, low-cost tools to perform MSI pre-screening. Here, the potential predictive role of the most relevant artificial intelligence-driven models in predicting microsatellite instability directly from histology alone is discussed, focusing on CRC. The role of deep learning (DL) models in identifying the MSI status is here analyzed in the most relevant studies reporting the development of algorithms trained to this end. The most important performance and the most relevant deficiencies are discussed for every AI method. The models proposed for algorithm sharing among multiple research and clinical centers, including federal learning (FL) and swarm learning (SL), are reported. According to all the studies reported here, AI models are valuable tools for predicting MSI status on WSI alone in CRC. The use of digitized H&E-stained sections and a trained algorithm allow the extraction of relevant molecular information, such as MSI status, in a short time and at a low cost. The possible advantages related to introducing DL methods in routine surgical pathology are underlined here, and the acceleration of the digital transformation of pathology departments and services is recommended.

Is Routine Pathology Evaluation of Tissue Removed for Fertility Preservation Necessary?

BackgroundFor all fertility preservation (FP) cases at our institution, a biopsy is performed for routine pathology from all gonadal tissue removed. This is not standard at all centers. We reviewed our experience with biopsy for pathological evaluation of ovarian and testicular specimens in FP cases to determine clinical utility. MethodsThe medical records of individuals who underwent ovarian tissue cryopreservation (OTC) or testicular tissue cryopreservation (TTC) between 2011 and 2023 were retrospectively reviewed under an IRB-approved study at a free-standing tertiary care children's hospital. Patient demographics, diagnosis, operative characteristics, and pathology results were collected. ResultsOne-hundred and eighty-three patients underwent OTC, and 134 patients underwent TTC. All patients had their gonadal tissue biopsied for routine pathology. Malignancy was identified in the biopsies of 4 OTC patients (2.2%) and 2 TTC patients (1.5%). Two OTC patients (1.1%) and 2 TTC patients (1.5%) did not have germ cells identified in their biopsy. All OTC and TTC patients and families elected to continue storing tissue for FP after discussion of pathology findings. ConclusionsPathology results provide another data point to help inform patients and their families when making decisions on ovarian or testicular tissue storage and on how tissue may be utilized in the future to restore fertility and/or hormones. There is a low rate of identifying malignancy in gonadal tissue biopsies taken from FP specimens even in patients with known malignancy. However, when malignancy was identified, it could be unexpected and alter the diagnosis and treatment plan significantly for patients. Level of EvidenceIV.

Mutational profiling of 103 unresectable pancreatic ductal adenocarcinomas using EUS-guided fine-needle biopsy.

Pancreatic ductal adenocarcinoma (PDAC) is among the most lethal cancers, with a 5-year survival rate of around 9%. Only 20% are candidates for surgery. Most unresectable patients undergo EUS-guided fine-needle biopsy (EUS-FNB) for diagnosis. Identification of targetable mutations using next-generation sequencing (NGS) is increasingly requested. Data on feasibility of EUS-FNB for NGS and knowledge regarding mutational profile of unresectable PDAC are scarce. We evaluated the "technical yield" of EUS-FNB for NGS in unresectable PDAC: relative fraction of diagnostic EUS-FNBs meeting technical criteria. We also investigated the "molecular yield": relative fraction of EUS-FNBs included in NGS containing sufficient DNA for detection of at least one mutation. Furthermore, we determined the relative frequency of cancer-associated mutations in unresectable PDAC. Formalin-fixed and paraffin-embedded EUS-FNBs diagnostic of unresectable PDAC and fulfilling these criteria were included (n = 105): minimum 3-mm2 tissue, minimum of 2-mm2 tumor area, and minimum 20% relative tumor area. NGS was performed using Ion GeneStudio S5 Prime System and Oncomine™ Comprehensive Assay v.3 including 161 cancer-related genes. Technical yield was 48% (105/219) and molecular yield was 98% (103/105). Most frequently mutated genes were KRAS (89.3%) and TP53 (69.9%), followed by CDKN2A (24.3%), ARID1A (9.7%), SMAD4 (7.8%), TSC2 (7.8%), and CCND3 (6.8%). EUS-FNB for NGS of unresectable PDAC is feasible. Our technical criteria for NGS, using leftovers in formalin-fixed and paraffin-embedded blocks after routine pathology diagnosis, were met by around half of EUS-FNBs. Almost all EUS-FNBs fulfilling the technical criteria yielded a successful NGS analysis.

(004) ROUTINE PATHOLOGY EXAMINATION OF EXCISED SPECIMEN FROM PATIENTS WITH NEUROPROLIFERATIVE VESTIBULODYNIA

Abstract Introduction Neuroproliferative vestibulodynia (NPV) is suspected in individuals with vulvas with complaints of entrance dyspareunia, vestibular allodynia and hyperalgesia, after ruling out other causes. When conservative biopsychosocial treatments for suspected NPV are unsuccessful, vestibulectomy may be an appropriate option. Excised vestibular surgical specimens are routinely sent for pathologic assessment including hematoxylin and eosin (H&amp;E) staining. Objective To review reports of gross and microscopic pathology examinations of excised vestibular specimens from a cohort of patients with NPV and compare immunohistochemical staining density among subgroups classified by severity of subepithelial inflammatory infiltrate. Methods Excised specimens from vestibulectomy surgeries were placed in 10% neutral buffered formalin prior to pathology examination. Routine pathology assessment included gross examination and cytological characterization of H&amp;E-stained vestibular tissue for malignancy, viral nuclear changes (in addition to P16 staining as needed), and quality of the vestibular epithelium and subepithelial inflammatory infiltrate. Severity of this infiltrate was classified by the pathologist as mild, mild-moderate, moderate, or severe. Additional vestibular tissue sections were processed for immunohistochemical staining for CD117 and PGP9.5, protein markers consistent with mast cells and nerves, respectively. Separating the tissues into subgroups based on severity of subepithelial inflammatory infiltrate, they were compared to mast cell counts and immunopositive mean fractional area for CD117 and PGP9.5 using the Kruskal-Wallis test. Results Gross examinations of the 1:00-11:00 (n = 63) and 12:00 (n = 54) regions of vestibular tissue showed mean dimensions (length, width, thickness) to be 5.8 x 2.1 x 0.6 cm and 1.0 x 0.7 x 0.3 cm, respectively. Microscopic examinations revealed no malignant cells in any specimen, and viral nuclear findings were consistent with HPV in 3 specimens. Vestibular tissue was characterized as non-reactive squamous epithelium in 91% and 93% of specimens for the 1:00-11:00 and 12:00 regions, respectively. Mature lymphocytes were found in the subepithelial stromal infiltrate of 67% of specimens in both the 1:00-11:00 and 12:00 regions. Severity of subepithelial stromal infiltrate was reported as: mild in 67% and 65%; mild-moderate in 5% and 4%; moderate in 14% and 15%; and severe in 14% and 17% of cases for the 1:00-11:00 and 12:00 regions, respectively. Among the subgroups categorized by severity of inflammatory infiltrate, there were no significant differences for CD117-immunopositive cell counts or mean fractional area of CD117 and PGP9.5 in the 1:00-11:00 and 12:00 regions. Conclusions Routine pathology examination including H&amp;E staining is standard for vestibulectomy specimens. There were no significant differences found among subgroups comparing severity of subepithelial inflammatory infiltrate with immunohistochemical data. Thus, information gleaned from routine pathology examination did not provide confirmation of the diagnosis of NPV. To better understand NPV, excised vestibular tissue should also be immunostained for CD117 and PGP9.5. Disclosure No.

A deep-learning-based model for assessment of autoimmune hepatitis from histology: AI(H).

Histological assessment of autoimmune hepatitis (AIH) is challenging. As one of the possible results of these challenges, nonclassical features such as bile-duct injury stays understudied in AIH. We aim to develop a deep learning tool (artificial intelligence for autoimmune hepatitis [AI(H)]) that analyzes the liver biopsies and provides reproducible, quantifiable, and interpretable results directly from routine pathology slides. A total of 123 pre-treatment liver biopsies, whole-slide images with confirmed AIH diagnosis from the archives of the Institute of Pathology at University Hospital Basel, were used to train several convolutional neural network models in the Aiforia artificial intelligence (AI) platform. The performance of AI models was evaluated on independent test set slides against pathologist's manual annotations. The AI models were 99.4%, 88.0%, 83.9%, 81.7%, and 79.2% accurate (ratios of correct predictions) for tissue detection, liver microanatomy, necroinflammation features, bile duct damage detection, and portal inflammation detection, respectively, on hematoxylin and eosin-stained slides. Additionally, the immune cells model could detect and classify different immune cells (lymphocyte, plasma cell, macrophage, eosinophil, and neutrophil) with 72.4% accuracy. On Sirius red-stained slides, the test accuracies were 99.4%, 94.0%, and 87.6% for tissue detection, liver microanatomy, and fibrosis detection, respectively. Additionally, AI(H) showed bile duct injury in 81 AIH cases (68.6%). The AI models were found to be accurate and efficient in predicting various morphological components of AIH biopsies. The computational analysis of biopsy slides provides detailed spatial and density data of immune cells in AIH landscape, which is difficult by manual counting. AI(H) can aid in improving the reproducibility of AIH biopsy assessment and bring new descriptive and quantitative aspects to AIH histology.

Prognostic role of tumor budding in oral squamous cell carcinoma: a retrospective study

Background Tumor budding is defined as a single cell or a cluster of up to five tumor cells at the invasion front of all oral squamous cell carcinoma (OSCC) cases. Tumor budding is considered as a marker of many important events in oral carcinoma, including the epithelial-mesenchymal transition, invasion, metastasis, and for further prognosis. Methods A total of 30 clinically and histopathologically diagnosed cases of OSCC were retrieved from the archival of the Department of Oral Pathology and Microbiology. After assessing the sections for the presence of tumor budding (TB), it is further categorized into high-intensity TB and low-intensity TB. Results We found a strong association between lymph node metastasis and histopathological grading in the current study. There was no association between TB and a patient’s survival rate. Conclusions This study emphasizes the significance of tumor budding, its relevance to prognosis, and reproducibility, as well as the significance of its evaluation and incorporation into routine surgical pathology practice in the management of OSCC. As a result, we believe that tumor budding is a crucial factor in determining tumor behavior.

Prediction of Epidermal Growth Factor Receptor Mutation Subtypes in Non–Small Cell Lung Cancer From Hematoxylin and Eosin–Stained Slides Using Deep Learning

Accurate assessment of epidermal growth factor receptor (EGFR) mutation status and subtype is critical for the treatment of non–small cell lung cancer patients. Conventional molecular testing methods for detecting EGFR mutations have limitations. In this study, an artificial intelligence–powered deep learning framework was developed for the weakly supervised prediction of EGFR mutations in non–small cell lung cancer from hematoxylin and eosin–stained histopathology whole-slide images. The study cohort was partitioned into training and validation subsets. Foreground regions containing tumor tissue were extracted from whole-slide images. A convolutional neural network employing a contrastive learning paradigm was implemented to extract patch-level morphologic features. These features were aggregated using a vision transformer-based model to predict EGFR mutation status and classify patient cases. The established prediction model was validated on unseen data sets. In internal validation with a cohort from the University of Science and Technology of China (n = 172), the model achieved patient-level areas under the receiver-operating characteristic curve (AUCs) of 0.927 and 0.907, sensitivities of 81.6% and 83.3%, and specificities of 93.0% and 92.3%, for surgical resection and biopsy specimens, respectively, in EGFR mutation subtype prediction. External validation with cohorts from the Second Affiliated Hospital of Anhui Medical University and the First Affiliated Hospital of Wannan Medical College (n = 193) yielded patient-level AUCs of 0.849 and 0.867, sensitivities of 79.2% and 80.7%, and specificities of 91.7% and 90.7% for surgical and biopsy specimens, respectively. Further validation with The Cancer Genome Atlas data set (n = 81) showed an AUC of 0.861, a sensitivity of 84.6%, and a specificity of 90.5%. Deep learning solutions demonstrate potential advantages for automated, noninvasive, fast, cost-effective, and accurate inference of EGFR alterations from histomorphology. Integration of such artificial intelligence frameworks into routine digital pathology workflows could augment existing molecular testing pipelines.

Electrical lymph node scanning (ELS) system for real-time intra-operative detection of involved axillary lymph nodes in adjuvant breast cancer patients

Lymph node (LN) status is an essential prognostic factor in breast cancer (BC) patients, with an important role in the surgical and therapeutic plan. Recently, we have been developed a novel system for real-time intra-operative electrical LN scanning in BC patients. The ELS scores were calibrated by pathological evaluation of the LNs. Herein, we evaluated the efficacy of ELS in a prospective study for non-chemo-treated breast cancer patients. This is a prospective study in which ELS scores are blind for pathologists who declare the clearance or involvement of LNs based on permanent pathology as the gold standard. ELS and frozen-section (FS) pathology results were achieved intra-operatively, and samples were sent for the permanent pathology. The score of ELS did not affect the surgeons’ decision, and the treatment approach was carried out based on FS pathology and pre-surgical data, such as imaging and probable biopsies. Patients were recruited from October 2021 through November 2022, and 381 lymph nodes of 97 patients were included in the study. In this study we recruited 38 patients (39.2%) with sentinel lymph node biopsy (SLNB) and 59 patients (60.8%) with ALND. Of the 381 LNs scored by ELS, 329 sentinel LNs underwent routine pathology, while others (n = 52) underwent both FS and permanent pathology. ELS showed a sensitivity of 91.4% for node-positive patients, decreasing to 84.8% when considering all LNs. Using ROC analysis, ELS diagnosis showed a significant AUC of 0.878 in relation to the permanent pathology gold standard. Comparison of ELS diagnosis for different tumor types and LN sizes demonstrated no significant differences, while increasing LN size correlated with enhanced ELS sensitivity. This study confirmed ELS’s efficacy in real-time lymph node detection among non-chemo-treated breast cancer patients. The use of ELS’s pathological scoring for intra-operative LN diagnosis, especially in the absence of FS pathology or for non-sentinel LN involvement, could improve prognosis and reduce complications by minimizing unnecessary dissection.

Ensemble-based deep learning improves detection of invasive breast cancer in routine histopathology images

Accurate detection of invasive breast cancer (IC) can provide decision support to pathologists as well as improve downstream computational analyses, where detection of IC is a first step. Tissue containing IC is characterized by the presence of specific morphological features, which can be learned by convolutional neural networks (CNN). Here, we compare the use of a single CNN model versus an ensemble of several base models with the same CNN architecture, and we evaluate prediction performance as well as variability across ensemble based model predictions.Two in-house datasets comprising 587 whole slide images (WSI) are used to train an ensemble of ten InceptionV3 models whose consensus is used to determine the presence of IC. A novel visualisation strategy was developed to communicate ensemble agreement spatially. Performance was evaluated in an internal test set with 118 WSIs, and in an additional external dataset (TCGA breast cancer) with 157 WSI.We observed that the ensemble-based strategy outperformed the single CNN-model alternative with respect to accuracy on tile level in 89 % of all WSIs in the test set. The overall accuracy was 0.92 (DICE coefficient, 0.90) for the ensemble model, and 0.85 (DICE coefficient, 0.83) for the single CNN alternative in the internal test set. For TCGA the ensemble outperformed the single CNN in 96.8 % of the WSI, with an accuracy of 0.87 (DICE coefficient 0.89), the single model provides an accuracy of 0.75 (DICE coefficient 0.78).The results suggest that an ensemble-based modeling strategy for breast cancer invasive cancer detection consistently outperforms the conventional single model alternative. Furthermore, visualisation of the ensemble agreement and confusion areas provide direct visual interpretation of the results. High performing cancer detection can provide decision support in the routine pathology setting as well as facilitate downstream computational analyses.